Establishing dynamic virtual roadway loop detectors with lane-level vehicle trajectory data

Recently there has been a significant expansion of roadway monitoring systems in many cities around the world. These systems are critical to traffic monitoring personnel as well as for drivers to get information on the roadway network conditions. Most of these systems are based on fixed embedded loop sensors. However, there are often gaps in these sensor networks or some areas don´t have sensors in place at all. To supplement these fixed sensor networks, it is also possible to obtain traffic information from probe vehicles. To date, many probe vehicle studies have provided data that were limited to link travel time estimates and rough estimates of traffic speed. As a result, they have had only limited penetration in an overall monitoring system. In this paper, we describe a low-cost system that is capable of lane-level positioning accuracy. By obtaining probe vehicle trajectories at the lane-level, it is possible to establish a true dynamic virtual roadway loop detector. These virtual sensors can be used to supplement the overall traffic network monitoring system. The developed system has been tested on a number of roadways and has performed very well when used with lane-level accurate roadway network data.